Anal. Chem. 1000, 62, 602-609
602
Adduct Ion Formation in the Methane-Enhanced Negative Chemical Ionization Mass Spectrometry of 2-(Alkylthio)- and 2-Alkoxy-s -triazines M. J. Incorvia M a t t i n a a n d Lee Q.Huang* Department of Analytical Chemistry, The Connecticut Agricultural Experiment Station, P.O. Box 1106, New Haven, Connecticut 06504 Roger N. Hayes Midwest Center for Mass Spectrometry, University of Nebraska, Lincoln, Nebraska 68588
+
+
Adduct Ions correspondlng to [M 111-, [M la]-, and [M 251- are observed at dgnHicant abundance In the highpressure, methanesnhanced, negative chemical ionlzatkm (NCI) mass spectra of seven 2-(alkytthb>s-trlazineszbUnder res. the same expetbnental comlitkns an adduct bn at [M 121’Is noted in the mass spectra of six 2-alkoxy-s-trlazlnes. These adduct Ions are interferences In the mass spectral analysis of these compounds. Detailed studles using methane-d4, “02/CH4, and Ar as alternate NCI enhancement gases reveal that the specks responslMe for adduct ion formaUon orlglnate from the methane gas. TMs was confirmed by h l g h - r e ” accurate mass measwoments of the adduct Ions. Colllslonally actlvated decomposition tandem mass spectrometry experiments were performed In an attempt to elucidate the structure of the adducts. The varlatlon of the adduct Ion abundance as a function of sample pressure, emlsskn cwent, repeller vdtage, and Ion w c e temperature Is also reported and possible mechanisms for the formatlon of the adduct Ions are examined.
+
+
INTRODUCTION The conditions prevailing in the high-pressure chemical ionization source of a conventional mass spectrometer are such that, in addition to unimolecular reactions (e.g., ionization of the analyte molecule and fragmentation of the parent ion), bimolecular, collisional reactions may occur with some facility ( I ) , producing ions that complicate the mass spectral analysis. Numerous examples of such collisional reactions have been reported for compounds whose spectra are acquired under electron capture negative chemical ionization (NCI) conditions (2-11). Examples from the positive chemical ionization (PCI) mode are also known (12). The reactions may be divided into two broad classes: class 1 reactions occur exclusively in the gas phase, while reactions belonging to class 2 appear to be mediated by the ion source walls. Class 1reactions include the gas-phase addition of a radical species, derived from the enhancement gas, to the analyte molecule to form a neutral adduct (2). The adduct may then be ionized by electron capture for detection in the negative ion mode or it may be protonated for detection in the positive ion mode. The PCI and NCI mass spectra of known radical traps such as 7,7,8&tetracyanoquinodimethane (TCNQ), tetracyanoethylene (TCNE), and tetracyanopyrazine (TCP) (3-61, as well as some PAH compounds (7,8), contain ions attributed to radical/molecule reactions. It has been shown that such gas-phase radical/molecule reactions in the high
* Author t o whom correspondence should be addressed.
pressure source occur at rates competitive with ionization (9). A second type of class 1reaction is an ion/molecule reaction in which both adduct formation and ionization occur in a single step. A well-known example is the reaction of CzH5+ from a hydrocarbon enhancement gas with analyte molecules to form the commonly detected [M + C,H,]+ adduct in PCI (1). Alternatively, both the fragment ion and the neutral molecule may be derived from the sample under analysis, giving rise to either anion adducts ( 1 0 , I I ) or cation adducts (12). Such gas-phase reactions are presumed to be independent of the moderating enhancement gas. Gas-phase ion/molecule reactions have also been reported for which the ionic component, such as 0‘-, is derived from enhancement gas impurities (13, 14). In contrast to class 1 reactions that occur solely in the gas phase, class 2 reactions are mediated by the source walls. Proposed wall-mediated reactions include the oxidation of PAH compounds by residual oxygen adsorbed on the ion source walls (15). The oxidized product is then desorbed from the source walls and ionized in the gas phase to produce unusual ions in the mas8 spectrum, some of which are observed a t m/z greater than the parent ion. In another example of a wall-mediated reaction carboxylic acids have been shown to react with adsorbed chlorine to form an adduct, which is subsequently desorbed and ionized in the gas phase (16). Because of their widespread use as herbicides, considerable interest exists in developing sensitive and specific analytical methods for s-triazine compounds by mass spectrometry using electron impact (EI) ionization (17, 18), PCI (19),and NCI (20,21). In our initial investigation of the utility of NCI for analysis of these compounds, several types of adduct ions at significant abundance were observed (20). In contrast to most gas-phase adducts reported previously for other classes of compounds, an s-triazine adduct may be observed at an abundance exceeding that of the parent ion and may appear at an mlz value corresponding to the parent ion of a different s-triazine. It is, therefore, apparent that such adducts are serious potential interferences in the accurate interpretation of the NCI mass spectra of s-triazines and merit a thorough investigation as to their origin and structure. The studies reported here include the use of argon, methane-d,, and W 2 / C H 4as alternate CI enhancement gases and identify the hydrocarbon enhancement gas as the source of the reactive species that lead to adduct formation. Collisionally activated decomposition (CAD) tandem mass spectrometry (MS/MS) is used in an attempt to elucidate the structures of the adduct ions. The data suggest that the alkylthio or alkoxy moiety in the 2-position of the s-triazine is an active site for reaction with the species derived from the enhancement gas. The abundance of the adduct ion as a function of the source temperature, the repeller voltage, the
0003-2700/90/0362-0602$02.50/00 1990 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 62, NO. 6, MARCH 15, 1990
603
Table I. Relative Abundance8 of Adduct Ions Observed in the Methane-Enhanced Negative Chemical Ionization Mass Spectra of 2-(AIkylthio)-~-triazines H
H
R,-C#-R,
T S-CH, name
mol w t
R1
RZ
ametryne desmetryne metoprotryne prometryne simetryne terbutryne dipropetryneb
227
Et Me i-Pr i-Pr Et Et i-Pr
i-Pr i-Pr (CHz)zOCH3 i-Pr Et t-Bu i-Pr
213 271 241
213 241
255
[M + 111-
relative abundance," % [M + 131[M+ 251-
44
100
69
100 100 100 100 100
80
46 63 78 14
[M - 11-
28 30 55 33 34 42 100
"Spectra were acquired via GC introduction of sample and DIP introduction of methane gas at ion source temperature of 300
* DiDrODetrVne is a 2-(SCH9CHq)-s-triazine.
emission current, and the sample concentration is examined in an attempt to elucidate the pathway for adduct ion formation. EXPERIMENTAL SECTION Reagents. The s-triazines at 98% to 99% purity were obtained from the EPA Pesticides and Industrial Chemicals Repository (Research Triangle, NC) or from the manufacturer, Ciba-Geigy, Agricultural Division (Greensboro, NC) and were used without further purification. All solvents were Fisher Optima grade. A separate methanol solution was prepared for each s-triazine at 100 to 200 mg/L concentration. Ultrahigh purity methane gas (Linde Specialty Gases) was filtered through an OMI-1 indicating purifier (Supelco, Bellefonte, PA) and an Oxiclear disposable gas purifier (Labclear, Oakland, CA). Argon (ultrahigh purity) was also supplied by Linde. Methanedl at 99.8% enrichment was obtained from Cambridge Isotopes (Woburn, MA) and lsOz at 98% enrichment from MSD Isotopes (Kirkland,Quebec, Canada). These stable isotope gases were supplied in glass flasks fitted with glass break seals. Instrumentation. Experiments were performed on a Hewlett-Packard 5890 gas chromatograph (GC) coupled via a capillary direct interface to an HP 5988A quadrupole mass spectrometer (MS) modified for negative ion detection. The methanol solution of the s-triazine was injected in splitless mode into the glass-lied GC inlet held at 250 O C . The temperature program for the SPB-608 column (Supelco) was 90 O C for 1 min, ramped at 10 OC/min to 250 "C and held for 2 min. The GC/MS interface was held at 250 O C . Helium, filtered through an OMI-1 indicating purifier, was used as the carrier gas. The MS tuning protocol has been reported previously (20). Unless otherwise indicated, methane, methane-d4,and argon were introduced into the source through the heated GC/MS interface by means of a Porter Instrument Co. (Hatfield, PA) Model DFC-1400 flow controller. Source pressure was monitored with a Hastings Raydist (Hampton, VA) DV-24 vacuum gauge tube located at the GC/MS interface. For experiments in which '8Oz was doped into methane, both the oxygen-18 and the methane were introduced into the source via a modified calibration probe (HP part number 05985-20585). In these experiments source pressures were measured with a Hastings Raydist vacuum gauge tube fitted to the end of the calibration probe. This probe is inserted into the source through the direct insertion probe (DIP) inlet port which on the HP 5988A is oriented on the source housing directly opposite the GC inlet port. The methane line was connected to the calibration probe after flow control via the DFC-1400 flow controller. The glass neck of the oxygen flask was attached to the calibration probe with appropriate size Teflon ferrules and swage fittings. Oxygen flow was controlled with a Circleseal shut-off valve (HP part number 0101-0297) and a Nupro metering valve (HP part number 0101-0287). There are two major differences between this design for enhancement gas introduction and the usual enhancement gas introduction through the GC/MS interface. First, the gas
OC.
line is not heated and second, as noted above, the gas is introduced into the source directly opposite the inlet port for the introduction of the GC effluent into the source. These differences required that enhancement gas source pressure values for the optimization of the ionization of DFTPP be reestablished by using the previously described method (20). By use of this instrumental configuration for the introduction of the enhancement gas into the MS source, a substantial decrease in the MS sensitivity for samples introduced via the GC was observed. In doping oxygen-18 into methane our procedure was to set the methane pressure to approximately 90% of its optimum value. Oxygen-18 was then leaked into the methane stream until the pressure in the source achieved the full optimum pressure value. High-resolution accurate mass measurements and CAD MS/MS spectra were acquired on a Kratos MS50 triple analyzer mass spectrometer of EBE configuration (22). Ions were formed in a Kratos Mark IV chemical ionization source: source temperature, 150 "C; electron energy, 280 eV; emission current, 500 p k accelerating voltage, 8 kV. The indicated methane pressure was 1 X lo6 Torr, corresponding to an estimated source pressure of lo-' Torr. For the CAD measurements, helium was used in the collision cell at pressures that achieved a 50% attenuation of the parent ion beam. RESULTS AND DISCUSSION Adduct ions observed for the seven 2-(alkylthio)-s-triazines included in the present study are summarized in Table I. Under the experimental conditions given in the table the base peak is the [M 131- ion and the [M + 251- ion is observed a t greater than 40% relative abundance for six 2-(methylthio)-s-triazines. In the NCI mass spectrum of dipropetryne, a 2-(ethylthio)+triazine, the [M + 111- ion is the only adduct observed at significant abundance. The adduct ions observed for six 2-alkoxy-s-triazines included in the present study are given in Table 11. For all six of the compounds an adduct ion is observed in the methane-enhanced NCI mass spectrum at an m / t value corresponding to [M + 12]'-. Elemental Composition of the Adduct Ions. Since a substantial amount of the ion current in the NCI mass spectra of 2-(alkylthio)-s-triaines(20)is carried by the [M - 11-rather than the M'- ion, we hypothesized that the observed [M + 131- adducts are due to [M + 14 - HI- rather than to an esoteric combination of elemental additions and losses, resulting in a net gain of 13 amu over the parent molecule. We, therefore, focused our attention on [M + 141 as the species formed, which undergoes subsequent loss of hydrogen and ionization. In this empirical description of the adduct, no mechanism for adduct formation, loss of hydrogen, or ionization is implied. Previous reports in the literature have attributed [M + 141'
+
ANALYTICAL CHEMISTRY, VOL. 62, NO. 6, MARCH 15,
604
1990
Table 11. Relative Abundances of Adduct Ions Observed in the Methane-Enhanced Negative Chemical Ionization Mass Spectra of 2-Alkoxy-s-triazines
I
DESYETRYNE
2,?s
1:12
H
0-R 184
relative abundance," mol
%
name
wt
R1
[M + 121'- [ M - 11-
atratone prometone secbumetone simetone terbutone "dipropetone"*
211
i-Pr i-Pr s-Bu Et Et i-Pr
225 225 197
225 239
R2
R
Et
Me
i-Pr
Me
Et Et t-Bu i-Pr
Me Me Me
Et
5 10 8 7 6 12
/
I
100 100 100 100 100 100
+
'
li
I.
I,
d
, I 1
1
MASS/CHARGE
DESMETRYNE
100,
"The spectra were acquired via GC introduction of the compound at an ion source temperature of 250 "C. b"Dipropetone"is not a commercially available compound; thus, it is named in quotation marks. adduds either to a wall-mediated oxidation reaction producing [M 0 - 2H]'- ions (15)or to a gas-phase radical/molecule reaction producing [M CH2]'- (7-9). In order to determine whether oxidation is responsible for adduct formation in the 2-(alkylthio)-s-triazines,oxygen-18 was doped into the methane stream. Figure l a is the NCI mass spectrum acquired using I8O2/CH4as the enhancement gas of desmetryne introduced through the GC into the source held at 300 "C. When this spectrum is compared with that shown in Figure lb, which was acquired under the same conditions except that 100% CHI was used as the enhancement gas, no mass shift of the adduct ions is noted. Thus, under our experimental conditions oxygen incorporation is not involved in the formation of these adduct ions for the 2-(alkylthio)+triazines. Subsequently, we focused our attention on the second plausible explanation for the production of the [M + 141 adduct, namely, that of a radical/molecule reaction in which the radical is derived from the hydrocarbon enhancement gas. An attempt to elucidate this role by using methane-d, as the enhancement gas is illustrated by the CD4enhanced NCI mass spectrum of desmetryne in Figure IC.Under the experimental conditions we observe an apparent hydrogen/deuterium exchange attributable to the N-H hydrogens of the amino moiety. We have shown previously that hydrogen/deuterium exchange for the hydrogens on the alkylamino groups is unlikely (20). These data also suggest that alkyl substitution of the alkylthio group by CD3, accompanied by exchange of the N-H hydrogen for deuterium, may be occurring. Hydrogen/deuterium exchange occurring during high-pressure NCI mass spectrometry using ND, as the enhancement gas has been reported previously (23). Our experiment confirms that an interaction between the hydrocarbon enhancement gas and the s-triazine substrate is taking place in the highpressure chemical ionization source. Indeed, if methane is replaced by argon as the enhancement gas, the [M + 131- and [M + 251- adduct ions are no longer observed, as may be seen in Figure I d for desmetryne. Similar results are obtained for all the (alkylthio)-s-triazines examined. Figure 2a shows the methane-enhanced NCI mass spectrum of approximately 200 ng of prometryne introduced through the GC. There are abundant [M + 131- and [M + 251- adduct ions. These ions are not detected in the argonenhanced NCI spectrum (Figure 2b) acquired under the same experimental conditions and recorded at the same retention time (RT) as the spectrum shown in Figure 2a. The methane-enhanced NCI mass spectrum of approximately 200 ng
111
198 180
2-21
IM+13)--
I
W
238
.'
I
z 0
a m 198
165
/
/
+
M A S S I C HA R 0 E
DESMETRYNE
1 0°1
W
z
I
0
/>I2
231
I
Ii
(Y-lr-
tc)
a